]> git-server-git.apps.pok.os.sepia.ceph.com Git - ceph.git/commit
mgr/nfs: Do not ignore clusters from rados pool conf objects 58284/head
authorAnoop C S <anoopcs@cryptolab.net>
Thu, 11 Jul 2024 07:46:30 +0000 (13:16 +0530)
committerPonnuvel Palaniyappan <pponnuvel@gmail.com>
Thu, 25 Jul 2024 17:10:11 +0000 (18:10 +0100)
commit667309396b576ed9136ec4f03c6c1a1a259690dd
tree6c84eac1ec0ea6962fe6303acffb52c89b7c74e9
parentf405b82b7e41c6e22e5ea00fbedfc7e822ed70d6
mgr/nfs: Do not ignore clusters from rados pool conf objects

Early handling of NoOrchestrator exception via a2192c73e41e prevented us
from looking for namespaces containing conf objects in .nfs pool. This
would mean that we fail to recognize such namespaces as clusters thereby
prohibiting the creation of nfs exports. We could instead search for
such namespaces while handling the NoOrchestrator exception within
available_clusters() before returning the final list of clusters.

Fixes: https://tracker.ceph.com/issues/66800
Signed-off-by: Anoop C S <anoopcs@cryptolab.net>
Conflicts:
src/pybind/mgr/nfs/export.py
src/pybind/mgr/nfs/export.py
src/pybind/mgr/nfs/utils.py